During system performance degradation, pinpointing underlying causes can be challenging, especially with datasets containing numerous features.
Phoenix leverages UMAP to visualize high-dimensional data during periods of performance degradation, thereby enabling the identification of clusters of problematic data.
